Overlay compensation method and system, semiconductor manufacturing method and device and storage medium

Abstract

The invention discloses an overlay compensation method and system, a semiconductor manufacturing method and device, and a storage medium. The overlay compensation method comprises the steps of obtaining to-be-compensated deviation data of a wafer subjected to photoetching processing; based on the to-be-compensated deviation data, a first overlay compensation model is adopted for fitting calculation so as to determine first model parameters of the first overlay compensation model, and the first overlay compensation model is a thin plate spline model; and based on the first overlay compensation model after the first model parameter is determined, obtaining a correction parameter for overlay compensation. According to the overlay compensation method disclosed by the embodiment of the invention, the thin-plate spline model is adopted, so that the production cost is reduced while the precision of overlay compensation is improved.

Description

Technical Field

[0001] Embodiments of the present invention relate to the field of semiconductor manufacturing, and in particular, to an overlay compensation method and system, a semiconductor manufacturing method, equipment, and a storage medium. Background Art

[0002] With the development of integrated circuit technology, the feature sizes to be realized on a wafer are getting smaller and smaller. Therefore, the lithography technology has higher and higher requirements for the lithography process and the accuracy of the lithography system.

[0003] Among them, in the lithography process, the pattern formed in the photoresist after exposure and development is usually referred to as the current layer pattern, and the pattern formed before this is usually referred to as the previous layer pattern. Overlay refers to the offset between the current layer pattern and the previous layer pattern, which is used to characterize the deviation of the relative position between the current layer pattern and the previous layer pattern. Since integrated circuit manufacturing usually requires forming multiple stacked film layers, if the current layer pattern and the previous layer pattern are not aligned, it will greatly affect the yield of the chip.

[0004] In the lithography process, in order to minimize the overlay as much as possible, after obtaining the overlay measurement data, the overlay measurement data will be analyzed, and the analysis result will be fed back to the alignment system of the lithography machine, so as to make further corrections for the next exposure to achieve overlay compensation.

[0005] However, it is difficult to balance the accuracy of overlay compensation and the production cost at present. Summary of the Invention

[0006] The problem solved by the embodiments of the present invention is to provide an overlay compensation method and system, a semiconductor manufacturing method, equipment, and a storage medium, which can improve the accuracy of overlay compensation while reducing the production cost.

[0007] To solve the above problems, an overlay compensation method provided by an embodiment of the present invention includes: obtaining the deviation data to be compensated of a wafer that has undergone lithography processing; based on the deviation data to be compensated, performing fitting calculation using a first overlay compensation model to determine the first model parameter of the first overlay compensation model, and the first overlay compensation model is a thin plate spline model ； Based on the first overlay compensation model after determining the first model parameter, obtaining the correction parameter for performing overlay compensation.

[0008] Correspondingly, an embodiment of the present invention further provides a semiconductor manufacturing method, including: performing lithography processing on a wafer to be compensated based on the correction parameter obtained by the overlay compensation method provided by the embodiment of the present invention, and realizing overlay compensation in the lithography processing.

[0009] Correspondingly, an embodiment of the present invention further provides a overlay compensation system, including: a to-be-compensated deviation data acquisition module, configured to acquire to-be-compensated deviation data of a wafer that has undergone lithography processing; a first model parameter acquisition module, configured to perform fitting calculation based on the to-be-compensated deviation data by using a first overlay compensation model to determine a first model parameter of the first overlay compensation model, where the first overlay compensation model is a thin-plate spline model; and a correction parameter acquisition module, configured to acquire a correction parameter for performing overlay compensation based on the first overlay compensation model after determining the first model parameter.

[0010] Correspondingly, an embodiment of the present invention further provides an electronic device, including at least one memory and at least one processor, where the memory stores one or more computer instructions, and one or more computer instructions are executed by the processor to implement the overlay compensation method provided by the embodiment of the present invention.

[0011] Correspondingly, an embodiment of the present invention further provides a storage medium, storing one or more computer instructions, where the one or more computer instructions are used to implement the overlay compensation method provided by the embodiment of the present invention.

[0012] Compared with the prior art, the technical solution of the embodiment of the present invention has the following advantages:

[0013] In the overlay compensation method provided by the embodiment of the present invention, after acquiring the to-be-compensated deviation data of the wafer that has undergone lithography processing, fitting calculation is performed based on the to-be-compensated deviation data by using a first overlay compensation model to determine a first model parameter of the first overlay compensation model, and the first overlay compensation model is a thin-plate spline (TPS) model. Compared with the solution of using a traditional polynomial function model as the overlay compensation model for overlay compensation, the thin-plate spline model has higher fitting accuracy and good generalization ability. Correspondingly, the requirement for the number of measurement points (i.e., overlay marks) during fitting calculation is lower. Therefore, on the one hand, this can reduce the requirement for the number of measurement points in the overlay compensation method, which not only reduces the process cost required for preparing overlay marks but also reduces the process cost required for measuring overlay deviation, thereby reducing the production cost. On the other hand, by using the thin-plate spline model, the probability of introducing noise data during the fitting process is lower, and it has strong convergence. Therefore, when performing fitting calculation based on the to-be-compensated deviation data by using the first overlay compensation model, it is beneficial to reduce the probability of occurrence of the over-fitting problem, improve the accuracy of the first model parameter, and correspondingly improve the accuracy of the correction parameter for performing overlay compensation, thereby improving the accuracy of overlay compensation. For this reason, the overlay compensation method of the embodiment of the present invention uses the thin-plate spline model, which is beneficial to reducing the production cost while improving the accuracy of overlay compensation.

[0014] In an alternative solution, before performing fitting calculation using the first set of overlay compensation models, fitting calculation is first performed using the second set of overlay compensation models. The second set of overlay compensation models is an m-order lens model based on a polynomial function, and the independent variable of the polynomial function is the coordinate information along the direction perpendicular to the exposure scanning direction within the exposure window (i.e., the exposure image field). Subsequently, based on the correction residuals corresponding to the second set of overlay compensation models, fitting calculation is performed using the first set of overlay compensation models. The parameters in the overlay compensation models usually have physical meanings. When the lens item needs to be considered in the overlay compensation model, the allowable distortion degree of the lens item usually has a large difference from other items. Therefore, two sets of overlay compensation models are separately set for the lens item, which facilitates the fitting calculation of the first set of overlay compensation models, is conducive to improving the accuracy of the first model parameters, and thus is conducive to improving the accuracy of the correction parameters used for overlay compensation. Moreover, the exposure window usually performs exposure along a certain exposure scanning direction. If it is considered that the contribution of the exposure window to the overlay deviation is fixed, then along the exposure scanning direction, the exposure-related data (e.g., light intensity, aberration, etc.) of the same field point in the exposure window at each position on the scanning path is basically the same. Therefore, by making the independent variable of the polynomial function corresponding to the second set of overlay compensation models be the coordinate information along the direction perpendicular to its exposure scanning direction (i.e., the X direction of the position coordinate system of the exposure field), the second set of overlay compensation models is a polynomial function that changes with the coordinate information perpendicular to its exposure scanning direction and has no component in the exposure scanning direction. Therefore, the process of fitting calculation is only to solve the model perpendicular to its exposure scanning direction, which is conducive to reducing the difficulty of solving and significantly reducing the probability of overfitting problems, making the accuracy of the calculated second model parameters relatively high, and correspondingly improving the accuracy of the correction parameters used for overlay compensation. In summary, by adopting a cascade model composed of the second set of overlay compensation models and the first set of overlay compensation models, the accuracy of overlay compensation is further improved.

[0015] In the semiconductor manufacturing method provided by an embodiment of the present invention, based on the correction parameters obtained by the overlay compensation method provided by the embodiment of the present invention, photolithography processing is performed on the wafer to be compensated, and overlay compensation is achieved during the photolithography processing. Since the overlay compensation method provided by the embodiment of the present invention can improve the accuracy of overlay compensation while reducing production costs, the overlay accuracy after photolithography processing based on overlay compensation is improved, thereby improving the product yield, and the production cost of the manufacturing method is reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 is a schematic flowchart of an embodiment of the overlay compensation method of the present invention;

[0017] Figure 2 It is a residual deviation distribution diagram after successively adopting the second set of lithography compensation model and the first set of lithography compensation model;

[0018] Figure 3 It is a schematic diagram of an embodiment of the lithography compensation system of the present invention;

[0019] Figure 4 It is a hardware structure diagram of an electronic device according to an embodiment of the present invention. Detailed implementation manners

[0020] As can be seen from the background art, it is difficult to balance the accuracy of lithography compensation and the production cost at present.

[0021] Since the linear compensation model is not sufficient to achieve lithography compensation, high order wafer alignment (HOWA) needs to be introduced. Generally, the more complex the compensation model is, the more parts it can correct. In current lithography compensation, a high order model of a traditional polynomial function is usually adopted. For example, a high order process correction (HOPC) model in the intra-field, that is, the iHOPC model.

[0022] After introducing high order correction, there are many parameters for high order correction. If the original number of lithography marks is still maintained, overfitting is likely to occur during the fitting calculation. Therefore, in order to be able to adopt high order correction, more lithography marks need to be prepared on the wafer to provide more measurement data for calculation. For example, if a polynomial function needs to adopt more than 38 correction terms, more than a hundred measurement points are required to obtain a better compensation result.

[0023] However, preparing more lithography marks on the wafer (that is, increasing the number of lithography marks in each exposure field) will increase the cost of preparing lithography marks and the cost of measurement, thus greatly increasing the production cost. Moreover, increasing the number of lithography marks in each exposure field will also occupy the area of effective devices, thus affecting the design requirements of customers.

[0024] Therefore, there is an urgent need to provide a new lithography compensation method that can improve the accuracy of lithography compensation while reducing the requirements for the number of measurement points.

[0025] To solve the above technical problems, an embodiment of the present invention provides a lithography compensation method. Referring to Figure 1 , a flowchart of an embodiment of the lithography compensation method of the present invention is shown. The lithography compensation method of this embodiment includes the following basic steps:

[0026] Step S1: Obtain the deviation data to be compensated for the wafer that has undergone lithography processing.

[0027] Step S2: Based on the deviation data to be compensated, perform fitting calculations using the first lithography compensation model to determine the first model parameters of the first lithography compensation model. The first lithography compensation model is a thin plate spline model.

[0028] Step S3: Based on the first lithography compensation model after determining the first model parameters, obtain the correction parameters for performing lithography compensation.

[0029] The first lithography compensation model is a thin plate spline. The first lithography compensation model has a high fitting accuracy and good generalization ability. Correspondingly, the requirement for the number of measurement points (i.e., registration marks) during fitting calculations is relatively low. Therefore, this can reduce the requirement for the number of measurement points in the lithography compensation method, not only reducing the process cost required for preparing registration marks but also reducing the process cost required for measuring the registration deviation, thereby reducing the production cost. Moreover, by using a thin plate spline, the probability of introducing noise data during the fitting process is relatively low, and it has strong convergence. Therefore, when performing fitting calculations, the probability of overfitting problems can be reduced, which is beneficial to improving the accuracy of the first model parameters and correspondingly improving the accuracy of the correction parameters for performing lithography compensation, thereby improving the accuracy of lithography compensation. For this reason, the lithography compensation method in the embodiments of the present invention uses a thin plate spline model, which is beneficial to reducing the production cost while improving the accuracy of lithography compensation.

[0030] To make the above objects, features, and advantages of the present invention more obvious and understandable, the following will describe the specific embodiments of the present invention in detail with reference to the accompanying drawings.

[0031] Figure 1 is a flowchart of an embodiment of the lithography compensation method of the present invention.

[0032] Refer to Figure 1 , perform step S1 to obtain the deviation data to be compensated for the wafer that has undergone lithography processing.

[0033] The wafer that has undergone lithography processing refers to the wafer after exposure and development processing. By obtaining the deviation data to be compensated for the wafer that has undergone lithography processing, preparations are made for subsequent lithography compensation based on the deviation data to be compensated.

[0034] In this embodiment, the wafer that has undergone lithography processing includes a substrate, a front layer formed on the substrate, and a current layer located on the front layer. A front layer pattern is formed in the front layer, and a current layer pattern is formed in the current layer. The deviation data to be compensated is obtained based on the registration offset between the current layer pattern and the front layer pattern.

[0035] Specifically, the front layer is an etched pattern layer, and the current layer is a photoresist layer remaining on the wafer after exposure and development processes. The photoresist layer is used as a mask for subsequent processes. As an example, the material of the photoresist layer includes photoresist.

[0036] In this embodiment, the deviation data to be compensated is used to characterize the exposure correction data that still needs to be performed in order to align the alignment marks in the current layer with those in the front layer, or, to characterize the current offset situation between the alignment marks in the current layer and those in the front layer.

[0037] Correspondingly, in this embodiment, the deviation data to be compensated is an offset vector. The direction of the offset vector is the offset direction of the alignment mark in the current layer relative to the corresponding alignment mark in the front layer, and the vector magnitude of the offset vector is the offset magnitude of the alignment mark in the current layer relative to the corresponding alignment mark in the front layer.

[0038] In this embodiment, the wafer is divided into multiple exposure fields (also known as shots). In the field of semiconductor manufacturing technology, for the convenience of the process, the wafer is usually divided into multiple exposure fields, which are arranged periodically and repetitively on the wafer. The exposure field is usually used as a basic exposure unit.

[0039] One or more alignment marks are formed in each exposure field, and the alignment marks are used as measurement points for overlay deviation measurement. Correspondingly, the deviation data to be compensated is obtained based on the alignment marks in the selected exposure field.

[0040] It should be noted that the number of alignment marks in each exposure field should be reasonable, or the number of alignment marks selected for overlay deviation measurement should be reasonable, so as to facilitate improving the accuracy of subsequent overlay compensation.

[0041] In this embodiment, obtaining the deviation data to be compensated for the wafer that has undergone the lithography process includes: obtaining the overlay deviation measurement data of the wafer that has undergone the lithography process; based on the overlay deviation measurement data, performing fitting calculations using a linear compensation model to determine the linear correction parameters of the linear compensation model; obtaining the correction residual corresponding to the linear compensation model after determining the linear correction parameters as the deviation data to be compensated.

[0042] It should be noted that the methods for obtaining the overlay deviation measurement data of the wafer that has undergone the lithography process include, but are not limited to, the Image Based Overlay (IBO) or the Diffraction - Based Overlay (DBO) for overlay measurement technology.

[0043] It should also be noted that the corrected residual value corresponding to the linear compensation model is the non-correctable error (NCE) data after correction by the linear compensation model. The corrected residual value is used to characterize the effectiveness of the overlay compensation. The smaller the corrected residual value, the better the overlay compensation effect.

[0044] Traditional overlay compensation methods first perform compensation through a linear compensation model, and then perform compensation through a high-order model based on a polynomial function. In this embodiment, first perform fitting calculation through a linear compensation model, and then combine with a thin plate spline model, so as to reduce the modification to the existing overlay compensation method.

[0045] Specifically, based on the overlay deviation measurement data, performing fitting calculation using a linear compensation model includes: fitting the linear compensation model according to multiple overlay deviation measurement data and the position coordinates (X, Y) of the measurement points corresponding to the multiple overlay deviation measurement data, and solving the linear correction parameters of the linear compensation model.

[0046] As an example, according to multiple overlay deviation measurement data and the position coordinates (X, Y) of the measurement points corresponding to the multiple overlay deviation measurement data, perform fitting calculation on the linear compensation model by the least squares method, and solve the linear correction parameters of the linear compensation model.

[0047] In other embodiments, a gradient algorithm can also be used for fitting calculation to solve the linear correction parameters of the linear compensation model.

[0048] Specifically, obtaining the corrected residual value corresponding to the linear compensation model after determining the linear correction parameters as the deviation data to be compensated includes: substituting the position coordinates (X, Y) of the measurement points corresponding to multiple overlay deviation measurement data into the linear compensation model to obtain overlay deviation calculation data; based on the difference between the overlay deviation measurement data and the overlay deviation calculation data, obtaining the corrected residual value corresponding to the linear compensation model.

[0049] It should be noted that if the linear compensation model is used for overlay compensation, it is usually difficult to compensate all measurement points. Therefore, after using the linear compensation model for overlay compensation, there will still be a corrected residual value. The corrected residual value corresponding to the linear compensation model is the remaining deviation after using the linear compensation model for overlay compensation.

[0050] In this embodiment, the overlay compensation method is suitable for performing on each exposure field on the wafer that has undergone lithography processing one by one. That is to say, after obtaining the correction parameters for performing overlay compensation, perform overlay compensation on each exposure field.

[0051] As an example, obtaining the deviation data to be compensated for a wafer that has undergone a lithography process includes: obtaining the overlay error measurement data for each exposure field on the wafer that has undergone a lithography process; obtaining the average value of the overlay error measurement data for multiple exposure fields; and obtaining the deviation data to be compensated based on the average value.

[0052] By using the average value of the overlay error measurement data for multiple exposure fields to obtain the correction parameters for overlay compensation, the complexity of the overlay compensation method is simplified.

[0053] In other embodiments, obtaining the deviation data to be compensated for a wafer that has undergone a lithography process includes: obtaining the deviation data to be compensated within any exposure field on the wafer that is to be compensated for overlay.

[0054] That is to say, after measuring the overlay error for each exposure field on the wafer, the correction parameters for each exposure field are calculated according to the corresponding measurement data, so that the lithography machine can use the correction parameters corresponding to each exposure field for each exposure field respectively.

[0055] The principle of overlay compensation is that it is considered that the overlay offset between the layer pattern and the previous layer pattern can be described by a mathematical model. However, in actual situations, the reasons for overlay errors are diverse, including the aberration of the lithography machine's optical system, the stability of the mask stage, etc. For different positions on the same wafer, it is difficult to accurately describe them only through a set of limited mathematical parameters. Therefore, each exposure field can be independently corrected to achieve separate compensation for each exposure field (correction per exposure, CPE).

[0056] By obtaining the deviation data to be compensated within the exposure field, it is convenient to independently correct each exposure field.

[0057] Independently correcting each exposure field can perform targeted compensation according to the overlay error situation of each exposure field, thereby further improving the accuracy of overlay compensation for the entire wafer.

[0058] Correspondingly, after subsequently obtaining the correction parameters for overlay compensation, the lithography machine uses the correction parameters corresponding to each exposure field for each exposure field.

[0059] Execute step S2. Based on the deviation data to be compensated, perform fitting calculations using the first overlay compensation model to determine the first model parameters of the first overlay compensation model, where the first overlay compensation model is a thin plate spline model.

[0060] The thin plate spline model has a high fitting accuracy and good generalization ability. Correspondingly, the requirement for the number of measurement points (i.e., registration marks) during fitting calculation is relatively low. Therefore, on the one hand, this can reduce the requirement for the number of measurement points in the registration compensation method, not only reducing the process cost required for preparing the registration marks but also reducing the process cost required for measuring the registration deviation, thereby reducing the production cost. On the other hand, by adopting the thin plate spline model, the probability of introducing noise data during the fitting process is relatively low, and it has strong convergence. Therefore, when performing fitting calculation based on the deviation data to be compensated and using the first registration compensation model, the probability of overfitting problems can be reduced, which is beneficial to improving the accuracy of the first model parameters and correspondingly improving the accuracy of the correction parameters used for registration compensation, thereby improving the accuracy of registration compensation.

[0061] In summary, the registration compensation method of this embodiment adopts the thin plate spline model, which is beneficial to reducing the production cost while improving the accuracy of registration compensation.

[0062] Therefore, the parameter information included in the TPS model includes offset vectors, position coordinates, and distortion parameters.

[0063] Specifically, the TPS model is a relationship model of offset vectors, position coordinates, and distortion parameters.

[0064] In this embodiment, the TPS algorithm is used for fitting calculation to solve the first model parameters of the first registration compensation model, and the first model parameters are used as the correction parameters for subsequent registration compensation.

[0065] Continue to refer to Figure 1 , in this embodiment, before using the first registration compensation model for fitting calculation to determine the first model parameters of the first registration compensation model, the registration compensation method further includes: performing step S4, based on the deviation data to be compensated, using the second registration compensation model for fitting calculation to determine the second model parameters of the second registration compensation model. The second registration compensation model is an m-order lens model based on a polynomial function, and the independent variable of the polynomial function is the coordinate information along the direction perpendicular to the exposure scanning direction within the exposure window.

[0066] The parameters in the overlay compensation model usually have physical meanings. When the lens term needs to be considered in the overlay compensation model, if the lens term is added to the first overlay compensation model, the allowable distortion degree of the lens term usually has a large difference from other terms. Therefore, a second overlay compensation model is separately set for the lens term. First, the second overlay compensation model is used for fitting calculation. Then, when the first overlay compensation model is used for fitting calculation later, the lens term does not need to be considered, which facilitates the fitting calculation of the first overlay compensation model, is beneficial to improving the accuracy of the first model parameters, and further beneficial to improving the accuracy of the correction parameters used for overlay compensation.

[0067] Moreover, the exposure window usually performs exposure along a certain exposure scanning direction. If it is considered that the contribution of the exposure window to the overlay deviation is fixed, then along the exposure scanning direction, the exposure-related data (such as light intensity, aberration, etc.) of the same field point in the exposure window at each position on the scanning path is basically the same. Therefore, by making the independent variable of the polynomial function corresponding to the second overlay compensation model be the coordinate information along the direction perpendicular to its exposure scanning direction (i.e., the X direction of the position coordinate system of the exposure field) within the exposure window, the second overlay compensation model is a polynomial function that changes with the coordinate information perpendicular to its exposure scanning direction and has no component in the exposure scanning direction. Therefore, the fitting calculation process is only to solve the model perpendicular to its exposure scanning direction, which is beneficial to reducing the difficulty of solving and significantly reducing the probability of overfitting problems, making the calculated accuracy of the second model parameters relatively high, and correspondingly improving the accuracy of the correction parameters used for overlay compensation.

[0068] In this embodiment, the second overlay compensation model is a high-order model (i.e., m is a natural number greater than 1). The higher complexity of the high-order model is beneficial to improving the accuracy of overlay compensation.

[0069] It should be noted that increasing the value of m is beneficial to improving the accuracy of overlay compensation. However, when the value of m is too large, it is also easy to cause the need for more data for fitting calculation, resulting in poor effect of reducing production costs.

[0070] Therefore, in this embodiment, in the m-order lens model, the value of m is 3, 4, or 5.

[0071] In this embodiment, the second overlay compensation model is a model based on the traditional Polynomial function.

[0072] That is to say, the second overlay compensation model is: the relationship model between the measured X-direction overlay deviation measurement data and the position coordinates (X, Y) at the (X, Y) position within the exposure field, and the relationship model between the measured Y-direction overlay deviation measurement data and the position coordinates (X, Y) at the (X, Y) position within the exposure field.

[0073] It can be understood that m in the m-th order represents the highest power of the X coordinate and the Y coordinate in the second set of lithography compensation models.

[0074] Since there is no component in the exposure scanning direction, when solving a model perpendicular to the exposure scanning direction, the requirement for the number of measurement points is relatively low. Therefore, even if the Polynomial function is used, the probability of overfitting is relatively low.

[0075] As an example, the exposure window is a slit. Correspondingly, the independent variable of the polynomial function is the coordinate information of the slit along its length direction, and the length direction of the slit is perpendicular to the exposure scanning direction.

[0076] In this embodiment, in the fitting calculation using the second set of lithography compensation models based on the deviation data to be compensated, the least squares method is used for fitting calculation to solve the second model parameters of the second set of lithography compensation models.

[0077] In other embodiments, a gradient algorithm can also be used for fitting calculation to solve the second model parameters of the second set of lithography compensation models.

[0078] For this reason, in this embodiment, the first set of lithography compensation models is used for fitting calculation to determine the first model parameters of the first set of lithography compensation models, including: obtaining the corrected residual corresponding to the second set of lithography compensation models after determining the second model parameters; based on the corrected residual corresponding to the second set of lithography compensation models, using the first set of lithography compensation models for fitting calculation to determine the first model parameters of the first set of lithography compensation models.

[0079] It should be noted that the second set of lithography compensation models is a lens model. Therefore, after performing lithography compensation using the second set of lithography compensation models, it is usually difficult to compensate all measurement points. Therefore, after performing lithography compensation using the second set of lithography compensation models, there will still be a corrected residual. The corrected residual corresponding to the second set of lithography compensation models is the remaining deviation after performing lithography compensation using the second set of lithography compensation models. In other words, the corrected residual corresponding to the second set of lithography compensation models is the NCE after being corrected by the second set of lithography compensation models.

[0080] Execute step S3 to obtain the correction parameters for performing lithography compensation based on the first set of lithography compensation models after determining the first model parameters.

[0081] After obtaining the correction parameters, during the subsequent photolithography process of the wafer to be compensated, lithography compensation is realized based on the correction parameters. That is to say, after correction based on the correction parameters, the subsequent wafer to be compensated is then subjected to photolithography processing.

[0082] Specifically, the correction parameters include the first model parameters obtained after fitting calculation of the first set of overlay compensation models.

[0083] In this embodiment, since the second set of overlay compensation models is used for fitting calculation before using the first set of overlay compensation models for fitting calculation, that is, the second set of overlay compensation models also provides part of the correction parameters. Therefore, among the correction parameters for overlay compensation, based on the first set of overlay compensation models after determining the first model parameters and the second set of overlay compensation models after determining the second model parameters, the correction parameters for overlay compensation are obtained.

[0084] That is to say, the correction parameters include the first model parameters obtained after fitting calculation of the first set of overlay compensation models and the second model parameters obtained after fitting calculation of the second set of overlay compensation models.

[0085] Therefore, by adopting a cascade model composed of the second set of overlay compensation models and the first set of overlay compensation models, the accuracy of overlay compensation is further improved.

[0086] Combined with reference Figure 2 , Figure 2 (a) is the distribution diagram of the remaining deviation amount after adopting the second set of overlay compensation models, Figure 2 (b) is the distribution diagram of the remaining deviation amount after adopting the first set of overlay compensation models. The numerical values of the remaining deviation amounts indicated by yellow are smaller than those indicated by red and blue. And the remaining deviation amounts indicated by yellow are relatively small. The remaining deviation amounts close to red are positive values, and the remaining deviation amounts close to blue are negative values. And the lighter the color of red, the smaller the numerical value of the remaining deviation amount (that is, closer to 0), the darker the color of red, the larger the numerical value of the remaining deviation amount. The lighter the color of blue, the smaller the numerical value of the remaining deviation amount, the darker the color of blue, the larger the numerical value of the remaining deviation amount. That is to say, along the direction from the deepest blue position to the deepest red position, the remaining deviation amount gradually changes from negative to positive.

[0087] In Figure 2 (a), the circled area is the yellow area, and the rest of the area is the blue area.

[0088] In Figure 2 (b), the circled area is the blue area, and the rest of the area is the yellow area. The remaining deviation amount of the yellow area is smaller than that of the blue area.

[0089] Combined with Figure 2 (a) and Figure 2(b) It can be seen that after first performing overlay compensation using the second overlay compensation model, the overlay accuracy within the exposure field is improved. After then performing overlay compensation using the first overlay compensation model, the overlay accuracy within the exposure field is further improved, and the distribution of the remaining deviation amount is more uniform.

[0090] Continuing to refer to Figure 1 , in this embodiment, after obtaining the correction parameters for performing overlay compensation, the overlay compensation method further includes: performing step S5 to perform an overlay deviation compensation feedback operation.

[0091] By performing the overlay deviation compensation feedback operation, the correction parameters are fed back to the lithography machine, so as to correct the alignment system of the lithography machine according to the correction parameters, generate corrected exposure parameters, and perform exposure processing based on the corrected exposure parameters, so that a higher alignment accuracy can be obtained after lithography processing of the wafer to be compensated.

[0092] For example, after obtaining the correction parameters for performing overlay compensation, an advanced process control (APC) system can be used to perform the overlay deviation compensation feedback operation.

[0093] Correspondingly, an embodiment of the present invention further provides a semiconductor manufacturing method. The semiconductor manufacturing method performs lithography processing on a wafer to be compensated based on the correction parameters obtained by the overlay compensation method described in the foregoing embodiment, and realizes overlay compensation during the lithography processing.

[0094] It can be understood that the wafer to be compensated can be a new batch of wafers or a wafer on which lithography processing has been currently performed.

[0095] Since the overlay compensation method provided in the foregoing embodiment can improve the accuracy of overlay compensation while reducing production costs, the overlay accuracy after lithography processing based on overlay compensation is improved, thereby improving the product yield, and moreover, the production cost of the manufacturing method is reduced.

[0096] As an example, the wafer to be compensated is a new batch (lot) of wafers. Therefore, during the lithography processing of the wafer to be compensated, lithography processing is performed on a new batch of wafers that belong to the same lithography process as the wafers on which lithography processing has been performed.

[0097] Here, belonging to the same lithography process means that: the purpose of performing the lithography processing is the same, that is to say, it belongs to the lithography process of the same process step. For example, it is used for forming a mask for etching a gate material layer.

[0098] In some other embodiments, the wafer to be compensated may also be a wafer on which lithography processing has been performed. Correspondingly, performing lithography processing on the wafer to be compensated includes: performing a rework process of lithography on the wafer on which lithography processing has been performed, and achieving overlay compensation in the rework process.

[0099] Specifically, performing a rework process of lithography on the wafer on which lithography processing has been performed includes: removing the existing photoresist layer on the wafer on which lithography processing has been performed to become the wafer to be compensated; after removing the existing photoresist layer, performing lithography processing on the wafer to be compensated again based on the correction parameters.

[0100] Correspondingly, an embodiment of the present invention further provides an overlay compensation system.

[0101] Reference Figure 3 , Figure 3 is a schematic diagram of an embodiment of the overlay compensation system of the present invention.

[0102] The overlay compensation system of this embodiment includes: a to-be-compensated deviation data acquisition module 10, configured to acquire the to-be-compensated deviation data of the wafer on which lithography processing has been performed; a first model parameter acquisition module 20, configured to perform fitting calculation using a first overlay compensation model based on the to-be-compensated deviation data to determine the first model parameters of the first overlay compensation model, where the first overlay compensation model is a thin plate spline model; a correction parameter acquisition module 30, configured to acquire correction parameters for performing overlay compensation based on the first overlay compensation model after determining the first model parameters.

[0103] The to-be-compensated deviation data acquisition module 10 is configured to acquire the to-be-compensated deviation data of the wafer on which lithography processing has been performed.

[0104] By acquiring the to-be-compensated deviation data of the wafer on which lithography processing has been performed, preparations are made for subsequent overlay compensation based on the to-be-compensated deviation data.

[0105] In this embodiment, the wafer on which lithography processing has been performed includes a substrate, a front layer is formed on the substrate, and a current layer is located on the front layer. A front layer pattern is formed in the front layer, and a current layer pattern is formed in the current layer. The to-be-compensated deviation data is obtained based on the overlay offset between the current layer pattern and the front layer pattern.

[0106] Specifically, the front layer is an etched pattern layer, and the current layer is a photoresist layer remaining on the wafer after exposure and development processing. The photoresist layer is used as a mask for subsequent processes. As an example, the material of the photoresist layer includes photoresist.

[0107] In this embodiment, the deviation data to be compensated is used to represent: the exposure correction data that still needs to be performed in order to align the alignment marks in the current layer with the alignment marks in the previous layer, or, to represent: the current offset situation of the alignment marks in the current layer and the alignment marks in the previous layer.

[0108] Correspondingly, in this embodiment, the deviation data to be compensated is an offset vector. The direction of the offset vector is the offset direction of the alignment marks in the current layer relative to the corresponding alignment marks in the previous layer, and the vector magnitude of the offset vector is the offset magnitude of the alignment marks in the current layer relative to the corresponding alignment marks in the previous layer.

[0109] In this embodiment, the wafer is divided into multiple exposure fields. In the field of semiconductor manufacturing technology, for the convenience of the process, the wafer is usually divided into multiple exposure fields. The exposure fields are arranged periodically and repetitively on the wafer, and the exposure fields are usually used as basic exposure units.

[0110] One or more alignment marks are formed in each exposure field, and the alignment marks are used as measurement points for measuring the overlay deviation. Correspondingly, the deviation data to be compensated is obtained based on the alignment marks in the selected exposure field.

[0111] It should be noted that the number of alignment marks in each exposure field should be reasonable, or the number of alignment marks selected for measuring the overlay deviation should be reasonable, so as to facilitate improving the accuracy of subsequent overlay compensation.

[0112] In this embodiment, the deviation data acquisition module 10 to be compensated includes: an overlay deviation measurement data acquisition unit (not shown in the figure), which is used to acquire the overlay deviation measurement data of the wafer that has undergone the lithography process; a linear correction parameter acquisition unit (not shown in the figure), which is used to perform fitting calculations based on the overlay deviation measurement data by using a linear compensation model to determine the linear correction parameters of the linear compensation model; a deviation data acquisition unit to be compensated (not shown in the figure), which is used to acquire the correction residual corresponding to the linear compensation model after determining the linear correction parameters as the deviation data to be compensated.

[0113] It should be noted that the method for the overlay deviation measurement data acquisition unit to acquire the overlay deviation measurement data of the wafer that has undergone the lithography process includes, but is not limited to, the overlay measurement technology based on imaging and image recognition or the overlay measurement technology based on diffraction.

[0114] It should also be noted that the correction residual corresponding to the linear compensation model is the non-correctable deviation data (NCE) after correction by the linear compensation model. The correction residual is used to represent the effectiveness of the overlay compensation. The smaller the correction residual, the better the effect of the overlay compensation.

[0115] In the traditional overlay compensation method, compensation is first performed through a linear compensation model, and then through a high-order model based on a polynomial function. In this embodiment, fitting calculation is first performed through a linear compensation model, and then combined with a thin plate spline model, so as to reduce the modification to the existing overlay compensation method.

[0116] Specifically, the linear correction parameter acquisition unit is used to fit a linear compensation model according to multiple overlay deviation measurement data and the position coordinates (X, Y) of the measurement points corresponding to the multiple overlay deviation measurement data, and solve the linear correction parameters of the linear compensation model.

[0117] As an example, the linear correction parameter acquisition unit fits a linear compensation model by the least squares method according to multiple overlay deviation measurement data and the position coordinates (X, Y) of the measurement points corresponding to the multiple overlay deviation measurement data, and solves the linear correction parameters of the linear compensation model.

[0118] In other embodiments, a gradient algorithm can also be used for fitting calculation to solve the linear correction parameters of the linear compensation model.

[0119] Specifically, the deviation data acquisition unit to be compensated substitutes the position coordinates (X, Y) of the measurement points corresponding to multiple overlay deviation measurement data into the linear compensation model to obtain overlay deviation calculation data, and obtains the correction residual corresponding to the linear compensation model based on the difference between the overlay deviation measurement data and the overlay deviation calculation data.

[0120] It should be noted that if the linear compensation model is used for overlay compensation, it is usually difficult to compensate all measurement points. Therefore, after using the linear compensation model for overlay compensation, there will still be a correction residual. The correction residual corresponding to the linear compensation model is the remaining deviation after using the linear compensation model for overlay compensation.

[0121] In this embodiment, the overlay compensation method is suitable for performing on each exposure field on the wafer that has been subjected to lithography processing one by one. That is, after obtaining the correction parameters for overlay compensation, overlay compensation is performed on each exposure field.

[0122] As an example, the deviation data acquisition module 10 to be compensated further includes: a data processing unit (not shown in the figure), which is used to obtain the average value of the overlay deviation measurement data of multiple exposure fields on the wafer that has been subjected to lithography processing.

[0123] Correspondingly, the deviation data to be compensated acquisition module 10 obtains the deviation data to be compensated based on the average value. Specifically, the linear correction parameter acquisition unit in the deviation data to be compensated acquisition module 10 performs fitting calculation using a linear compensation model based on the average value.

[0124] The correction parameters for performing overlay compensation are obtained through the average value of the overlay deviation measurement data of multiple exposure fields, thereby simplifying the complexity of the overlay compensation method.

[0125] In other embodiments, the deviation data to be compensated acquisition module is used to obtain the deviation data to be compensated in any exposure field on the wafer that has undergone lithography processing and needs overlay compensation.

[0126] That is to say, after measuring the overlay deviation of each exposure field on the wafer, the correction parameters of each exposure field are calculated according to the corresponding measurement data, so that the lithography machine can use the correction parameters corresponding to each exposure field for each exposure field respectively.

[0127] The principle of overlay compensation is that it is considered that the overlay offset between the layer pattern and the previous layer pattern can be described by a mathematical model. However, in actual situations, the reasons for overlay errors are diverse, including the aberration of the lithography machine optical system, the stability of the mask stage, etc. For different positions on the same wafer, it is difficult to accurately describe them only through a set of limited mathematical parameters. Therefore, each exposure field can be independently corrected to achieve separate compensation for each exposure field.

[0128] By obtaining the deviation data to be compensated in the exposure field, it is convenient to independently correct each exposure field.

[0129] Independent correction of each exposure field can perform targeted compensation according to the overlay deviation situation of each exposure field, thereby further improving the accuracy of the overlay compensation for the entire wafer.

[0130] Correspondingly, after subsequently obtaining the correction parameters for performing overlay compensation, the lithography machine uses the correction parameters corresponding to each exposure field for each exposure field.

[0131] The first model parameter acquisition module 20 performs fitting calculations through the first overlay compensation model to determine the first model parameters of the first overlay compensation model. The first overlay compensation model is a thin plate spline model, which has high fitting accuracy and good generalization ability. Correspondingly, the requirement for the number of measurement points (i.e., overlay marks) during fitting calculations is relatively low. Therefore, on the one hand, this can reduce the requirement for the number of measurement points in the overlay compensation method, which not only reduces the process cost required for preparing overlay marks but also reduces the process cost required for measuring overlay deviation, thereby reducing the production cost. On the other hand, by using the thin plate spline model, the probability of introducing noise data during the fitting process is relatively low, and it has strong convergence. Therefore, when performing fitting calculations based on the deviation data to be compensated and using the first overlay compensation model, the probability of overfitting problems can be reduced, which is beneficial to improving the accuracy of the first model parameters and correspondingly improving the accuracy of the correction parameters used for overlay compensation, thereby improving the accuracy of overlay compensation.

[0132] In summary, the overlay compensation system of this embodiment uses a thin plate spline model, which is beneficial to reducing the production cost while improving the accuracy of overlay compensation.

[0133] Therefore, the parameter information included in the TPS model includes offset vectors, position coordinates, and distortion parameters.

[0134] Specifically, the TPS model is a relationship model of offset vectors, position coordinates, and distortion parameters.

[0135] In this embodiment, the first model parameter acquisition module 20 uses the TPS algorithm to perform fitting calculations and solve the first model parameters of the first overlay compensation model. The first model parameters are used for subsequent overlay compensation.

[0136] Continue to refer to Figure 3 , in this embodiment, the overlay compensation system further includes: a second model parameter acquisition module 40, which is used to perform fitting calculations based on the deviation data to be compensated using a second overlay compensation model to determine the second model parameters of the second overlay compensation model before the first model parameter acquisition module 20 determines the first model parameters of the first overlay compensation model. The second overlay compensation model is an m - order lens model based on a polynomial function, and the independent variable of the polynomial function is the coordinate information along the direction perpendicular to its exposure scanning direction within the exposure window.

[0137] The parameters in the overlay compensation model usually have physical meanings. When the lens term needs to be considered in the overlay compensation model, if the lens term is added to the first overlay compensation model, the allowable distortion degree of the lens term usually differs significantly from other terms. Therefore, a second overlay compensation model is separately set for the lens term. First, the second overlay compensation model is used for fitting calculation. Then, when the first overlay compensation model is used for fitting calculation subsequently, the lens term does not need to be considered, which facilitates the fitting calculation of the first overlay compensation model, is conducive to improving the accuracy of the first model parameters, and further conducive to improving the accuracy of the correction parameters used for overlay compensation.

[0138] Moreover, the exposure window usually performs exposure along a certain exposure scanning direction. If it is considered that the contribution of the exposure window to the overlay deviation is fixed, then along the exposure scanning direction, the exposure-related data (such as light intensity, aberration, etc.) of the same field point in the exposure window at each position on the scanning path is basically the same. Therefore, by making the independent variable of the polynomial function corresponding to the second overlay compensation model be the coordinate information along the direction perpendicular to its exposure scanning direction (i.e., the X direction of the position coordinate system of the exposure field) within the exposure window, the second overlay compensation model is a polynomial function that changes with the coordinate information perpendicular to its exposure scanning direction and has no component in the exposure scanning direction. Therefore, the fitting calculation process is only to solve the model perpendicular to its exposure scanning direction, which is conducive to reducing the difficulty of solving and significantly reducing the probability of overfitting problems, making the calculated accuracy of the second model parameters relatively high, and correspondingly improving the accuracy of the correction parameters used for overlay compensation.

[0139] In this embodiment, the second overlay compensation model is a high-order model (i.e., m is a natural number greater than 1). The higher complexity of the high-order model is conducive to improving the accuracy of overlay compensation.

[0140] It should be noted that increasing the value of m is conducive to improving the accuracy of overlay compensation, but when the value of m is too large, it is also easy to cause the need for more data for fitting calculation, resulting in poor effect of reducing production costs.

[0141] Therefore, in this embodiment, in the m-order lens model, the value of m is 3, 4, or 5.

[0142] In this embodiment, the second overlay compensation model is a model based on a traditional polynomial function.

[0143] Since there is no component in the exposure scanning direction, when solving a model perpendicular to the exposure scanning direction, the requirement for the number of measurement points is relatively low. Therefore, even if the Polynomial function is used, the probability of overfitting problems is relatively low.

[0144] As an example, the exposure window is a slit. Correspondingly, the independent variable of the polynomial function is the coordinate information of the slit along its length direction, and the length direction of the slit is perpendicular to the exposure scanning direction.

[0145] In this embodiment, the second model parameter obtaining module 40 performs fitting calculation by using a second lithography compensation model based on the deviation data to be compensated, and performs fitting calculation by using the least squares method to solve the second model parameters of the second lithography compensation model.

[0146] In other embodiments, the second model parameter obtaining module may also perform fitting calculation by using a gradient algorithm to solve the second model parameters of the second lithography compensation model.

[0147] For this reason, in this embodiment, the first model parameter obtaining module 20 includes: a correction residual obtaining unit (not shown in the figure), configured to obtain the correction residual corresponding to the second lithography compensation model after determining the second model parameters; a first model parameter obtaining unit (not shown in the figure), configured to perform fitting calculation by using a first lithography compensation model based on the correction residual corresponding to the second lithography compensation model to determine the first model parameters of the first lithography compensation model.

[0148] It should be noted that the second lithography compensation model is a lens model. Therefore, after performing lithography compensation by using the second lithography compensation model, it is usually difficult to compensate all measurement points. Therefore, after performing lithography compensation by using the second lithography compensation model, there will still be a correction residual. The correction residual corresponding to the second lithography compensation model is the remaining deviation after performing lithography compensation by using the second lithography compensation model. In other words, the correction residual corresponding to the second lithography compensation model is the NCE after being corrected by the second lithography compensation model.

[0149] After obtaining the correction parameters, during the subsequent lithography process of the wafer to be compensated, lithography compensation is implemented based on the correction parameters. That is to say, after correction based on the correction parameters, the subsequent wafer to be compensated is subjected to lithography processing.

[0150] Specifically, the correction parameters include the first model parameters obtained after performing fitting calculation on the first lithography compensation model.

[0151] In this embodiment, before using the first set of lithography compensation model for fitting calculation, the second set of lithography compensation model is first used for fitting calculation, that is, the second set of lithography compensation model also provides some correction parameters. Therefore, the correction parameter acquisition module 30 obtains the correction parameters for lithography compensation based on the first set of lithography compensation model after determining the first model parameters and the second set of lithography compensation model after determining the second model parameters. That is to say, the correction parameters include the first model parameters obtained by fitting the first set of lithography compensation model and the second model parameters obtained by fitting the second set of lithography compensation model.

[0152] Therefore, by adopting a cascade model composed of the second set of lithography compensation model and the first set of lithography compensation model, the accuracy of lithography compensation is further improved.

[0153] Continue to refer to Figure 3 , in this embodiment, the lithography compensation system further includes: a feedback module 50, configured to perform a lithography deviation compensation feedback operation after obtaining the correction parameters for lithography compensation.

[0154] By performing the lithography deviation compensation feedback operation, the correction parameters are fed back to the lithography machine, so that the alignment system of the lithography machine is corrected according to the correction parameters, corrected exposure parameters are generated, and exposure processing is performed based on the corrected exposure parameters, so that a higher alignment accuracy can be obtained after lithography processing of the wafer to be compensated.

[0155] For example, the feedback module 50 may be an advanced process control (APC) system.

[0156] It should be noted that, as an example, the lithography compensation system is used to execute the lithography compensation method provided in the foregoing embodiment. Therefore, for the specific description of the lithography compensation system, reference may be made to the relevant records in the foregoing embodiment.

[0157] Correspondingly, an embodiment of the present invention further provides an electronic device, which can implement the lithography compensation method provided in the embodiment of the present invention by loading the above lithography compensation method in the form of a program.

[0158] Refer to Figure 4 , which shows the hardware structure diagram of the electronic device provided in an embodiment of the present invention. The device in this embodiment includes: at least one processor 01, at least one communication interface 02, at least one memory 03, and at least one communication bus 04.

[0159] In this embodiment, the number of the processor 01, the communication interface 02, the memory 03, and the communication bus 04 is at least one, and the processor 01, the communication interface 02, and the memory 03 complete communication with each other through the communication bus 04.

[0160] The communication interface 02 may be an interface of a communication module for network communication, for example, an interface of a GSM module.

[0161] The processor 01 may be a central processing unit CPU, or a specific integrated circuit ASIC (Application Specific Integrated Circuit), or one or more integrated circuits configured to implement the overlay compensation method of this embodiment.

[0162] The memory 03 may include a high-speed RAM memory, and may also include a non-volatile memory, such as at least one disk memory. Among them, the memory 03 stores one or more computer instructions, and the one or more computer instructions are executed by the processor 01 to implement the overlay compensation method provided in the foregoing embodiment.

[0163] It should be noted that the above-mentioned implementation of the electronic device may further include other devices (not shown) that may not be necessary for the disclosure of the embodiments of the present invention; in view of the fact that these other devices may not be necessary for understanding the disclosure of the embodiments of the present invention, the embodiments of the present invention do not introduce them one by one.

[0164] The embodiments of the present invention further provide a storage medium, and the storage medium stores one or more computer instructions, and the one or more computer instructions are used to implement the overlay compensation method provided in the foregoing embodiment.

[0165] The above embodiments of the present invention are combinations of the elements and features of the present invention. Unless otherwise mentioned, the elements or features can be regarded as selective. Each element or feature can be practiced without being combined with other elements or features. In addition, the embodiments of the present invention can be constructed by combining some elements and / or features. The operation sequence described in the embodiments of the present invention can be rearranged. Some configurations of any embodiment can be included in another embodiment, and can be replaced by the corresponding configuration of another embodiment. It is obvious to those skilled in the art that the claims that do not have an explicit reference relationship with each other in the appended claims can be combined into the embodiments of the present invention, or can be included as new claims in the amendments after the submission of this application.

[0166] Embodiments of the present invention can be implemented by various means such as, for example, hardware, firmware, software, or a combination thereof. In a hardware configuration, the method according to an exemplary embodiment of the present invention can be implemented by one or more application specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), processors, controllers, microcontrollers, microprocessors, and the like.

[0167] In a firmware or software configuration, embodiments of the present invention can be implemented in the form of modules, procedures, functions, etc. The software code can be stored in a memory unit and executed by a processor. The memory unit is located inside or outside the processor and can send data to and receive data from the processor via various known means.

[0168] The foregoing description of the disclosed embodiments enables those skilled in the art to implement or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present invention. Thus, the present invention will not be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0169] Although the present invention is disclosed as above, the present invention is not limited thereto. Any person skilled in the art can make various changes and modifications without departing from the spirit and scope of the present invention. Therefore, the protection scope of the present invention should be determined by the scope defined in the claims.

Claims

1. A lithography overlay compensation method, characterized in that, it includes: obtaining the deviation data to be compensated of a wafer that has undergone lithography processing; based on the deviation data to be compensated, performing fitting calculations using a first lithography overlay compensation model to determine the first model parameters of the first lithography overlay compensation model, where the first lithography overlay compensation model is a thin plate spline model; based on the first lithography overlay compensation model after determining the first model parameters, obtaining correction parameters for performing lithography overlay compensation.

2. The lithography overlay compensation method according to claim 1, characterized in that, obtaining the deviation data to be compensated of a wafer that has undergone lithography processing includes: obtaining the lithography overlay deviation measurement data of a wafer that has undergone lithography processing; based on the lithography overlay deviation measurement data, performing fitting calculations using a linear compensation model to determine the linear correction parameters of the linear model; obtaining the correction residual corresponding to the linear compensation model after determining the linear correction parameters as the deviation data to be compensated.

3. The lithography overlay compensation method according to claim 1, characterized in that, before performing fitting calculations using a first lithography overlay compensation model to determine the first model parameters of the first lithography overlay compensation model, the lithography overlay compensation method further includes: based on the deviation data to be compensated, performing fitting calculations using a second lithography overlay compensation model to determine the second model parameters of the second lithography overlay compensation model, where the second lithography overlay compensation model is an m - order lens model based on a polynomial function, and the independent variable of the polynomial function is the coordinate information along the direction perpendicular to its exposure scanning direction within the exposure window; performing fitting calculations using a first lithography overlay compensation model to determine the first model parameters of the first lithography overlay compensation model includes: obtaining the correction residual corresponding to the second lithography overlay compensation model after determining the second model parameters; based on the correction residual corresponding to the second lithography overlay compensation model, performing fitting calculations using a first lithography overlay compensation model to determine the first model parameters of the first lithography overlay compensation model; in obtaining the correction parameters for performing lithography overlay compensation, based on the first lithography overlay compensation model after determining the first model parameters and the second lithography overlay compensation model after determining the second model parameters, obtaining the correction parameters for performing lithography overlay compensation.

4. The lithography overlay compensation method according to claim 3, characterized in that, in the m - order lens model, the value of m is 3, 4, or 5.

5. The lithography overlay compensation method according to claim 3, characterized in that, the exposure window is a slit, the independent variable of the polynomial function is the coordinate information along the length direction of the slit, and the length direction of the slit is perpendicular to the exposure scanning direction.

6. The lithography overlay compensation method according to claim 3, characterized in that, in performing fitting calculations using a second lithography overlay compensation model based on the deviation data to be compensated, fitting is performed by the least squares method.

7. The lithography overlay compensation method according to any one of claims 1 to 6, characterized in that, Obtaining the deviation data to be compensated for a wafer that has undergone lithography processing includes: obtaining the overlay error measurement data for each exposure field on the wafer that has undergone lithography processing; obtaining the average value of the overlay error measurement data for multiple exposure fields; and obtaining the deviation data to be compensated based on the average value. Alternatively, Obtaining the deviation data to be compensated for a wafer that has undergone lithography processing includes: obtaining the deviation data to be compensated within any one exposure field on the wafer that has undergone lithography processing.

8. The overlay compensation method according to claim 1, wherein, after obtaining the correction parameters for performing overlay compensation, the overlay compensation method further includes: performing an overlay error compensation feedback operation.

9. A semiconductor manufacturing method, wherein, it includes: Performing lithography processing on the wafer to be compensated based on the correction parameters obtained from the overlay compensation method according to any one of claims 1 to 8, and implementing overlay compensation during the lithography processing.

10. The semiconductor manufacturing method according to claim 9, wherein, during the lithography processing of the wafer to be compensated, lithography processing is performed on a new batch of wafers that belong to the same lithography process as the wafer that has undergone lithography processing.

11. The semiconductor manufacturing method according to claim 9, wherein, performing lithography processing on the wafer to be compensated includes: performing a rework process of lithography on the wafer that has undergone lithography processing, and implementing overlay compensation during the rework process.

12. The semiconductor manufacturing method according to claim 11, wherein, performing a rework process of lithography on the wafer that has undergone lithography processing includes: removing the existing photoresist layer on the wafer that has undergone lithography processing to become the wafer to be compensated; after removing the existing photoresist layer, performing lithography processing on the wafer to be compensated again based on the correction parameters.

13. An overlay compensation system, wherein, it includes: A deviation data acquisition module to be compensated, configured to acquire the deviation data to be compensated for a wafer that has undergone lithography processing; A first model parameter acquisition module, configured to perform fitting calculations using a first overlay compensation model based on the deviation data to be compensated to determine the first model parameters of the first overlay compensation model, and the first overlay compensation model is a thin plate spline model; A correction parameter acquisition module, configured to acquire the correction parameters for performing overlay compensation based on the first overlay compensation model after determining the first model parameters.

14. An electronic device, wherein, it includes at least one memory and at least one processor, and the memory stores one or more computer instructions, wherein the one or more computer instructions are executed by the processor to implement the overlay compensation method according to any one of claims 1 to 8.

15. A storage medium, wherein, the storage medium stores one or more computer instructions, and the one or more computer instructions are used to implement the overlay compensation method according to any one of claims 1 to 8.

Images